Optimal scale-invariant signal-to-noise ratio and curriculum learning for monaural multi-speaker speech separation in noisy environment.
Chao MaDongmei LiXupeng JiaPublished in: APSIPA (2020)
Keyphrases
- signal to noise ratio
- noisy environments
- speech enhancement
- noise reduction
- scale invariant
- speaker verification
- speech recognition
- speaker identification
- multiscale
- noisy speech
- natural images
- edge detection
- affine invariant
- audio visual
- speech signal
- automatic speech recognition
- scale space
- single channel
- color images
- computer vision
- keypoints
- object recognition
- minimum mean square error
- background noise
- image structure
- computer simulation
- non stationary